Comment Blog 13 March, 2024

The folly of mission-based industrial policy

Dr Bryan Cheang
Assistant Director of the Centre for the Study of Governance and Society, King's College London

Industrial policy is back on the agenda. Mariana Mazzucato argues that we ought to view governments as “investors, risk takers, and innovators”. The State should tackle grand social challenges in a mission-directed manner, just as the American Government once put a man on the moon.

However, Mazzucato’s arguments rest on a straw man of “market fundamentalism”, a belief that individuals “maximise his/her own objective function”, and in so doing lead to maximally efficiency outcomes for society.

While some neoclassical economists think that market outcomes do approximate the expectations of this model, the case for markets need not rest on these unrealistic theories.

A better approach is a process-oriented one, with the market seen as a “discovery process” whereby individuals — often acting in very imperfect circumstances — nonetheless discover opportunities and in so doing achieve a semblance of coordination with each other. The reason why markets are preferable to governments is because they facilitate the comparison of multiple products and services at one time, as compared to the necessarily singular decisions of governments. Put simply, when multiple experiments can be tried out at one go, the possibility of systemic error is reduced and trial-and-error learning can take place.

Therefore, while one would be right to reject Economics 101 this does not mean accepting Government 101, assuming that the constraints faced by private actors are either absent or minimised in the political process. Mazzucato is right to claim that market actors are “embedded in rules, norms and contracts affecting organisational behaviour”. But so too are political actors, who are embedded in a web of stakeholder relationships including interest groups, their constituents and other agencies. Such embeddedness could mean that states fail to exert sufficient capacity to drive the economy in a mission-directed manner.

Viewing the market as a discovery process of trial-and-error learning is not an esoteric academic exercise. It means taking seriously the sort of incentives and knowledge held by actors in both markets and politics.

The serious problem of Mazzucato’s position, relating to the aspect of knowledge, is defining the “right mission” to pursue. Advocates of mission-led governance typically use the Apollo and Manhattan projects as success stories, but these are cases which aimed to “develop a particular technological capability, and the achievement of their technological objective signalled the end of the program”. This problem is further clarified by Johan Larsson, who argues that “missions are most likely to achieve intended ends, when reasonable people agree on the problem, what needs to be done, and when responsibility can be assigned”.

The problem is that most recent proposals for mission-led governance pursue ambitious open-ended goals. In such cases, there is uncertainty of “what to do, where accountability is unassigned, and where the failure-success axis cannot be meaningfully assessed”.

Policy problems favoured by advocates of mission-led governance, like climate change, are complex, non-linear and wicked problems. There are multiple policy levers needed to “tackle climate change” involving multiple stakeholders, conflicting objectives and diverse trade-offs. What might work at one specific locale may not in others, a problem that is especially pertinent to economic innovation.

Advocates of industrial policy typically point to breakthrough inventions or products arising from state investment as proof of success. This begs the question of what are the opportunity costs? What were the alternative ideas, investments and technologies that were not put into practice? Governments typically struggle to answer this question because they lack the same counterfactuals that exist in the market sphere.

This is why, even though governments have produced many remarkable achievements in the past, they nonetheless cannot act entrepreneurially. By channelling resources to specific goals, governments have been successful in achieving inventive breakthroughs — for example the atomic bomb. But these are accomplishments of engineering, not entrepreneurship. The former is about achieving a specific end regardless of cost, while the latter is about transforming resources to create more value than they cost. This is not to say that governments cannot try to engage in entrepreneurship. They simply struggle to tell whether the resources they used were worth more than they cost.

In the absence of clear-cut evaluation metrics, political criteria start entering the calculus. When political criteria, objectives and aims take centre stage, the door to interest group influence and inefficiency is opened. It was shown, for example, that the Strategic Innovation Programs (SIP) in Sweden and the SHOK programs in Finland ended up favouring large established firms and universities, thus sustaining existing structures rather than paving the way for new innovations.

Similar inefficiency was detected in China’s wind power industrial policy. Significant investment and the construction of wind power plants resulted in substantial economic activity but limited value creation. The sector was deeply regulated by administrative practices and command-and-control targets, focusing more on installed capacity rather than actual electricity production. While domestic patents in the Chinese wind power sector were plentiful there were few international patents, indicating a lack of globally competitive technological innovation.

Advocates of industrial policy sometimes insist that these problems can be managed by organisational design. For example, through certain control mechanisms, one could curb rent-seeking pressures. Knowledge considerations could also be overcome by incorporating effective evaluation within bureaucracies, consultation with experts, and a “learning culture” amongst policymakers. On this front, scholars typically point to East Asian states as model “innovation bureaucracies” for others to emulate.

Yet, pointing to East Asia is problematic. Social conditions are drastically different. The relatively communitarian culture makes it easier for the State to enforce compliance by social actors to the chosen mission. Such a communitarian culture is also closely associated with the authoritarian powers that East Asian states have utilised over recent history.

Moreover, empirical evidence has shown that even Singapore — said to be the “most successful” with industrial policy — failed to foster a sustainable ecosystem of indigenous entrepreneurship. Its best efforts at controlling for rent-seeking through “embedded autonomy” have failed to prevent a culture of rent-seeking and subsidy entrepreneurship from flourishing.

These international experiences should act as a cautionary tale for those seeking mission-oriented governance and an entrepreneurial state.


Dr Bryan Cheang is the Assistant Director of the Centre for the Study of Governance and Society at King's College London. He is also a Research Fellow at STICERD, London School of Economics. His research is on comparative political economy, complexity and the efficacy of industrial policy. 

As part of the Reform Scholars programme he published 'Thinking outside the sandbox: how regulatory relief and learning can promote innovation'.